The present invention relates generally to protein purification employing chromatographic techniques. More particularly, the present invention relates to procedures for the rapid and efficient isolation of biologically active proteins such as glycoproteins and especially glycoproteins having high sialic acid content (e.g., erythropoietic factors such as erythropoietin) from natural (e.g., blood fractions and urine) and recombinant (e.g., genetically transformed mammalian cell culture fluid) sources.
Numerous techniques have in the past been applied in preparative separations of biochemically significant materials. Commonly employed preparative separatory techniques include: ultrafiltration, column electrofocusing, flat-bed electrofocusing, gel filtration, electrophoresis, isotachophoresis and various forms of chromatography. Among the commonly employed chromatographic techniques are ion exchange chromatography and adsorption chromatography. The former process is a separatory method wherein fluid components with differing net charges are distinguished and isolated by means of elution (stepwise or with a continuously applied gradient) with eluents of differing ionic strength. A gel matrix (resin) carrying either a positive or negative charge is employed to adsorb (bind) components of opposing net charge. During desorption (elution) charged sample components are exchanged by salt ions in the selected eluent, with specific sample components eluting at specific ionic strengths. Reverse phase adsorption chromatography involves separation of fluid sample components based on differing polarity. Sample components are adsorbed to a granulated gel matrix (resin) by non-covalent bonds. Thereafter, stepwise or continuous gradient elution results in, selective desorption of components upon exchange with a non-polar solvent in the eluent.
While the numerous separatory techniques mentioned above are routinely employed in the separation of relatively small hydrophobic and hydrophylic molecules, they have somewhat limited applicability in preparative separations of relatively large molecules such as proteins, especially complex proteins such as lipoproteins, nucleoproteins and glycoproteins. Illustrative of the state of the art in protein separations are reviews by Brown, et al., Analytical Biochemistry, 99, 1-21 (1979) and Rubinstein, Analytical Biochemistry, 99, 1-7 (1979). See also, "VYDAC.TM. Comprehensive Guide to Reverse Phase Materials for HPLC", The Sep/A/Ra/Tions Groups, Hesperia, Calif. and the publication of co-applicants Strickland and co-workers in Parsons, et al., Endocrinology, 114, (6), 2223-2227 (1984). Further, to the extent that, for example, reverse phase HPLC procedures have been suggested or employed in isolations of proteins or polypeptides, non-polar solvents generally recommended have included reagents that are difficult to handle or to separate from the desired protein such as acetonitrile. See Parsons, et al., supra. Only a single reference is known to exist disclosing elution with ethanol, specifically aqueous ethanol/formic acid mixtures. See Takagaki, et al. Journal of Biological Chemistry, 5, (4), 1536-1541 (1980).
The apparent limited utility of the abovenoted techniques in preparatory separations of high molecular weight complex proteins is especially problematic in view of recent intensive efforts directed toward isolation, purification and application to therapeutic, immunoprophylactic and diagnostic procedures of a wide variety of complex viral and eucaryotic proteins available in only minute quantities from natural sources wherein they are found in association with myriad other complex proteins. As one example, biochemically significant mammalian hematopoietic factors such as erythropoietin, thrombopoietin, granulopoietin and granulocytemarcophage colony stimulating factor are available in extremely small quantities from urine of aplastic anemia patients. Recovery procedures from urinary fluid sources have generally been very complex, costly and labor-intenstive and have generated relatively low yields of active product. A widely practiced method for obtaining biologically active preparations of urinary erythropoietin (a high molecular weight, high sialic acid content glycoprotein) may be found in Miyake, et al., Journal of Biological Chemistry, 252 (15), 5558-5564 (1979). The seven-step procedure includes ion exchange chromatography, ethanol precipitation, gel filtration, and adsorption chromatography and is reported to provide a 21% yield of glycoprotein with 70,400 Units/mg potency.
The extensive application of recombinant methodolgies to the large scale preparation of eucaryotic proteins has substantially enhanced the prospects for obtaining the desired molecules in quantity and in some instances even simplified purification procedures needed to obtain biologically active materials. Illustratively, where the desired recombinant proteins need not be glycosylated to possess biological activity, large quantities of protein can often be produced in E.coli recombinant hosts in the form of insoluble "inclusion bodies" which contain few proteinaceous contaminants, proteases, or the like. Where glycosylation and/or host membrane processing to develop proper secondary and tertiary conformation are required for biological activity, however, eucaryotic hosts such as yeast and mammalian cells in culture (e.g., COS-1 and CHO cells) provide more suitable recombinant hosts. Use of such hosts, however, generally gives rise to increased difficulty in recovery of biologically active forms of proteins in good yield. Host cell lysates frequently include proteinaceous constituents of sufficiently similar molecular weight, charge, polarity and solubility characteristics (vis-a-vis the recombinant protein) to make ready separation difficult. Further proteolytic enzymes endogenous to the host provide a relatively chronic source of biological activity loss for the desired protein. Where recombinant products are secreted into media supernatants by the host cells, similar problems attend isolation from, e.g., culture media from growth of transformed mammalian cell cultures owing principally to the complexity of the media employed.
There thus continues to exist a need in the art for rapid and efficient preparative separatory procedures suitable for recovery of biologically active proteins from fluid sources and most especially for recovery of complex recombinant proteins such as recombinant erythropoietin from variously "contaminated" fluids such as mammalian cell culture supernatants.
The disclosures of co-owned, co-pending U.S. patent application Ser. No. 675,298, entitled "Production of Erythropoietin", filed Nov. 30, 1984, by FuKuen Lin (corresponding to PCT No. US84/02021, filed Dec. 11, 1984, scheduled for publication June 20, 1985 as No. WO85/02610) are specifically incorporated by reference herein for the purpose of relating the background of the present invention, especially with respect to the state of the art regarding recombinant methodologies applied to large scale production of mammalian erythropoietin.